STOP
Early Journal Content on JSTOR, Free to Anyone in the World
This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in
the world byJSTOR.
Known as the Early Journal Content, this set of works include research articles, news, letters, and other
writings published in more than 200 of the oldest leading academic journals. The works date from the
mid-seventeenth to the early twentieth centuries.
We encourage people to read and share the Early Journal Content openly and to tell others that this
resource exists. People may post this content online or redistribute in any way for non-commercial
purposes.
Read more about Early Journal Content at http://about.istor.org/participate-istor/individuals/early-
journal-content .
JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people
discover, use, and build upon a wide range of content through a powerful research and teaching
platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit
organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please
contact support@jstor.org.
SOME VERB-FORMS IN GERMANIC
1. The Injunctive
The injunctive, which in form is an unaugmented indicative,
functions in part as a preterit : Skt. bhdrat, Gr. <^^p€ 'he bore' (cf.
Brugmann, Kz. vergl. Gram. 551). According to Dieter, Altgerm.
Dial. II, §§ 203, 7, 189, 2, this use of the injunctive in Germ, is
confined to such forms as ON. kom 'came' and the second sing. pret.
in WGerm. Failure to see the injunctive in the preterit of a whole
class of verbs is apparently due to a tenacious adherence to the old
theory of the so-called reduplicating verbs.
In these verbs the second and third series correspond most closely
to the regular ablaut-verbs. Here the regular preterit of the redu-
plicating verbs in N. and WG. have (or had originally) the same
vowel as the present of the regular ablaut-verbs. That is, they are
injunctives of the type Skt. bhdrat, Gr. <t>ip€. Of the same type are
several preterits with Germ, eu or e in regular ablaut-verbs of the
second and third series.
The following preterits with Germ, eu have either IE. eu or
eu, which would fall together: ON. bid, (bua 'wohnen'):Skt. bhdvati
'wird, ist,' OE. beon 'be.' OE. heof:heofan 'lament,' Goth, hiufan
'wehklagen.' OE. hneop 'plucked': Goth, dishniupan ' zerreissen.'
OE. onreod 'inbuit ':reorfon 'redden,' ON. rio'tSa, Gr. ipevdu. OE.
beat {beatan 'beat'): root *bheud- (Walde*, 185). OE. deog {deagiari
' dye') ideogol, deagol, dlegle 'hidden,' OHG. tougal 'dunkel, ver-
borgen,' tougan 'dunkel, finster, verborgen,' Lett, dukt 'matt sein'
(hinsinken); OHG. tuhhan 'tauchen,' petochen 'versunken,' fer-
tochenen 'verborgenen,' etc. (cf. Color-Names 89; IE. a^ 74 f.).
OHG. screot (scrotan 'abschneiden'):ON. skrio'Sr 'Fetzen, zerfetztes
Buch.' OHG. stioz (stozan 'stossen'):root *steud- 'tundere' (cf.
Walde^, 798f.).
The following preterits have Germ, e, i from IE. e or, in some
cases, e, which would be shortened before a liquid or nasal com-
bination: ON. giall (gall): gialla 'gellen,' OHG. gellan, etc. ON.
121] 57 [MoDEBH Philologt, June, 1916
58 Francis A. Wood
hialp (halp, OSwed. halpa ' helf en ') : ON. hialpa, Goth, hilpan ' helf en.'
ON. blett from *blind (blanda 'mischen'):Goth. blinds 'blind,'
Lith. blendzius 'sich verfinstern.' ON. fell, felt, feldom (falda 'den
Kopf bedecken/ Goth, falpan 'falten'):root *pel-t- 'fold.' ON.
gekk, gingom {ganga 'gehen'):3d sing. pres. gingr, Lith. iengiit,
'schreite.' ON. hell, helt, heldom (halda 'halten,' Goth, haldan
'weiden,' Gr. povKoKos 'Rinderhirt'):Gr. (cAojuai, /ceXXco (Uhlenbeck,
Et. Wb.^ 72). OS. skeld {skaldan 'schalten, stossen'): OHG. scel-
fan 'schelten.' OS. well {wallan 'wallen'): ON. vella 'sieden.' OS.
weld (waldan 'walten'):Lith. veldeti 'regieren, besitzen.'
Outside of the Norse the original Germ, e, i of the preterit is
changed by analogy of other types. Thus the pret. vowel in OHG.
becomes either ia, ie after the analogy of hiaz 'hiess,' liaz 'liess,' or
eo, io after the analogy of liof 'lief.' OHG. spialt for * spelt {spaltan
'spalten'):OE. speld 'splinter.' OHG. v}ialk for *welk {walkan
'walken'):Lith. vMi 'walken.' OHG. wiah for *welz (walzan
' walzen ') : ON. velta 'walzen.' OHG. spian for *spinn (spannan
'spannen'):spwnan 'spinnen,' root *spe- 'stretch.' This is practi-
cally the same as my original explanation in Germanic Studies II.
2. The Dttal of the Verb in Germanic
The Goth. 1st dual pret. witu exactly corresponds to Skt.
vidvd. In ON. this is represented by Run. waritu {*writu) 'wir
beide ritzten,' ON. toko vit 'we two took' (cf. Noreen Aisl. Gr.
§ 461, 3). This gives the best explanation for the apparent loss of
final -m in tdko vSr 'we took' as compared with ver tokom. That is,
for original vit toko, vir tokom and toko vit, tokom v&r arose vit, ver
tokom and toko vit, v6r. In some cases also the loss of final -m may
have been phonetic, as certainly was the case with -3 of the 2d plur.
before an initial />- (Noreen Aisl. Gr. § 459, 4).
In WG. the 1st dual of the pret. would likewise end in -w; OE.
aru, nutu {wit) we, etc. Forms ending in -a may represent Goth.
-OS as in nimos 'we two take': gonga, walla, wyrca we. The forms
in -e are evidently opt., as binde wit, we, bunde we, etc. : Goth, bin-
daiwa, *bundeiwa, 'WG. *bindeu, *bundlu. From these forms the
usage spread to the 2d dual and plur., as: binde git, ge, bunde ge,
etc. (cf. Sievers Ags. Gr. § 360).
122
Some Verb-Fokms in Germanic 59
OHG. urizzu in wizzuwir T 132, 17 is identical with Goth. witu.
Forms in -e are opt. : wege wir, sage wir, genese wir, etc. (cf . Braune
Ahd. Gr. § 307, Anm. 7). In the 1st plur. pres. the opt. forms early
crowded out the ind., hence the more frequent forms in -e. In
MHG. the forms without n where the verb was followed by wir
became more common.
The 2d dual in Goth, ends in -tsisaihwats, wituts, mleits, etc.
This regularly represents the IE. suffix -tes or -thes, which would
become in Germ. -3zz or -J>iz, according to accent. The i would dis-
appear in Goth., and there would result -J>s, which would be assim-
ilated to -ts. We may assume that Goth, -ps would always become
-ts when not prevented by association. Similarly -J>s becomes -ts,
z in ON., as: orz, gen. of or'd, kuazk from kiia'd sik, etc. (cf. Noreen,
Aisl. Gr. §183, d); and ts in OE., as: snltst 'du schneidest' for
snlpst (cf. Sievers Ags. Gr. § 359, Anm. 7).
In ONorw. the usual form in the 2d plur. ends in -ir, as bindir
'bindet.' Noreen, Pauls Grundriss ¥, 637 f., derives this from
*bindi(t-r from bindid ir 'bindet ihr.' But bindir may be more
simply and easily explained as coming from *bindid-r, Germ. *bindi-
"Siz, the 2d dual.
In WG. a *bindi'Siz (Lat. regitis) or *binda'Siz (Goth, bindats)
would fall together with the 2d plur.: OHG. nemet, nemat, more
rarely quidit, ferit for guedet, faret (cf. Braune Ahd. Gram. § 308).
The a in Goth, nimats is probably not original. But if so, then it is
identical with OHG. nemat. The pret. ind. nemuts, nemup and opt.
nemeits, nemeip fall together completely in OHG. ndmut, ndmlt.
Cf . Brugmann, Grundriss II, 2, p. 1373.
3. West Germanic deda 'Did'
For WG. deda 'did' the most satisfactory explanation is that
it is a reduplicated perfect. But whatever its ordinal form it is
evident that it has suffered a number of analogical changes. In
general it may be compared with Skt. dadhau (cf . Brugmann, Grdr.
IF, 3, p. 457), though not corresponding exactly, for this would give
OS. *dedo, OHG. *teto. But d^da does not stand in Germ, by itself.
It is like Goth, saiso, lailo, waiwo, ON. sera, rera (from sd 'saen,'
123
60 Fkancis a. Wood
r6a 'rudern'), and other reduplicated preterits modeled after them,
as: slera {sld 'schlagen'), grera (groa 'keimen'), snera (snwa' wenden'),
gnera {gnOa 'schaben'), bnere (Goth, hnauan 'zerreiben').
From these we can reconstruct, if not the original IE. forms, at
least the common Germ, forms from which they sprang. Of the
Goth, verbs the following forms are preserved: 1st and 3d sing.
saiso, 2d sing, saisost, 3d plur. lailoun, waiwoun. These forms are
no doubt nearest the Germ, forms, but they have all suffered an
analogical change. For the stem saiso- (for *saizd-, pre-Germ.
*ses6-) is carried through after the model of lailot. A pre-Germ. 1st
sing. *ses6 or *ses6m could only give Goth. *salza, and a 3d sing.
*ses6t would give the same. The form that exists {saiso) is plainly
analogical. The 2d sing, saisost is either made over from *saisos,
Germ. *sezos, with t tacked on from the usual ending of the 2d sing,
perf. as: wast, gaft, falht, etc., or else, which is less likely, changed
from *sezop{a), with st taking the place of the proper perf. ending,
after the analogy of ga-stost, wast, waist, laist, gast, etc. The plural
must have been in Germ.: *sezom^, *sezop{i), *sezon{p). These
were changed in Goth, by adding the usual endings of the pret.
plur. to saiso-, giving *saisoum, *saisoup, saisoun for *saisom, etc.
In ON. the corresponding sera is conjugated in the sing, like
a weak verb. In the plur. serom, sera's, sero can represent either
the original Germ, forms with o or the later analogical forms with u.
In OS. the forms of deda that correspond most closely to the
above are 1st and 3d sing, deda, -e, 2d sing, dedos (older *dedos),
plur. dedun, representing both the 1st pi. dedum (for older *dedom)
and the 3d pi. dedun (older *deddn). The 2d sing, dadi and pi.
dadun are plainly after the analogy of a strong verb of the 5th class.
The same is true of OHG. tati, tdtum, tatut, tdtun. But even in
OHG. the form *tetun must have occurred, being represented by
MHG. teten, and the opt. tete (:OHG. *teti, OS. dedi). Moreover,
there is evidence that the earUer forms for OHG. were tela, *tetos,
*tetom, *tetot, Heton. For this best explains the endings in Alem.
suohtom, suohtot, suohton, and the 2d sing. -ds{t) in all the dialects.
Compare OS. sohtos, -as ( : dedos) for the more common sohtes. The
e Ln OS. dedun also points to an original -om, -on, for -wm, -un would
have given *didun.
124
Some Verb-Forms in Germanic 61
4. The e- Aorist in Germanic and the Weak Preterit
Attention has been called to a number of weak preterits without
a dental. Examples are: Olcel. mis-gere, Run. kiari 'machte,' pi.
A;mrw:ON. gerua 'machen'; WN. horfe 'wandte,' horfo 'wandten
sich' : horf a 'sich wenden' {huerfa, Goth, hwairban, etc.); ODan.
havce, OSw. have, Olcel. hafe 'haA,te':hafa 'haben'; OSw. laghi,
ODan. laghcR 'legte,' OSw. laghu 'legten': Goth, lagjan 'legen';
ODan. saghoe, OSw. saghe (with a according to Noreen, Altschw.
Gram. § 553, A18), seghi 'sagte/ Olcel. sogod 'sagtet': OHG. sagen
'sagen'; OSw. leghi 'mietete': Olcel. leiga 'mieten,' Goth, leihwan
'leihen' (cf. Noreen, Pauls Grundriss P, 635; Aschw. Gr. §§ 552
Anm. 5, 553 Anm. 5, 9, 13, 18). Here also belong OHG. fermisson
'vermissten,' pret. opt. missm (cf. Braune, AM. Gr. §363 Anm. 6)
and Goth, iddja 'ging.'
Put into Goth, as forms of sagen this would give: *saga, sages,
saga; sagum, sagup, sagun. In ending, these correspond exactly
with the weak preterit. Thus the forms of WG. sagda would be the
same as the above with d after g. However, this means only that
the words were conjugated after a common tj^je. For the plural
the terminations are the same as in all preterits in Germ. They
therefore prove nothing as to the original form. The singular,
however, points to an aorist of the type of Gr. Hbb-p-qv, hhK-qv, etc.
The same formation also gives e-presents : Gr. iT&.K-i] ' schwand dahin' :
Lat. tacet, OHG. daget 'schwe^t'; Gr. ■^Xt^ij 'anointed': OHG.
Uhet 'lebt'; Gr. kjiavri ' raged': OHG. /ormonei 'verachtet,' etc.
Similarly we may compare ON. hafe 'hatte':OHG. Jialet 'hat';
ON. horje '-wandte' -.horf er 'kehrt um'; OSw. saghe 'sagte': OHG.
saget 'sagt'; ON. -gere, kiari 'machte' :0E. (Merc.) gearwep 'macht.'
Goth, iddjes can hardly be anything but an e- aorist with pre-
served augment: Germ. *ijjes from *e-ie-s : Goth, jer 'jahr,' MHG.
jan 'fortlaufende Reihe, Strich,' NHG. dial, jan 'gerader Strich,
Reihe, die der Maher einhalt,' Skt. y&na-hi, 'Bahn,' y&li 'geht,'
Czech jki 'fahren, reiten,' Lith. joti id., etc., base *ie-, *ia 'go, ire'
(cf. Schade 462). Here also probably OHG. jamar 'leidvoU,' sh.
'Schmerzgefilhl, das ein herber Verlust erzeugt, Herzeleid, schmerz-
liches Verlangen': Skt. yama 'Gang, Weg, Zug; das Angehen,
Anrufen, Flehen'; yOcati 'fleht, heischt, fordert, bittet um,' ya-
125
62 Francis A. Wood
'gehen, ziehen; gelangen zu, kommen nach, angehen um,' yapdyati
'macht gehen': Gr. Iij/jli 'let go, send away; set in motion, throw;
mid. go eagerly towards: long for, yearn after, wish, desire.'
The conjugation of iddja in pre-Germ. would therefore be:
*eiem, -es, -et; -eue, -etes; Smen, -ete, -int. This, with doubling
of the i, would give in Germ.: *ijjq, -es, -e; -ewi, -ediz; -em^, SSi,
-en.^ In the dual and plur. the forms were modeled in Goth, after
the perfect by adding the perfect endings to the stem ijjed-, which
was abstracted from the 2d dual and 2d plur., perhaps also influenced
by a Goth. *didum. In WG. ije- would naturally become by analogy,
at least in the plur., *ijo- or *iju-, whence OE. eo-de, ea-de, lo-de.
This gives the type for all the weak verbs in Gothic. That is,
to the preterit stem, as: iddj-, brdht-, mund-, kunp-, wiss-, nasid-,
etc., were added the terminations -a, -e-s, -a; [-e-du], -e-duts; -e-dum,
e-dup, -e-dun. Whatever may be the origin of these stems, in feeling
the forms must have been divided, e.g., for the 3d plur., iddje-dun,
brdhte-dun, munde-dun, kunpe-dun, wisse-dun, naside-dun, habaide-
dun, salbode-dun, etc. For otherwise the forms would be without
system, and Gothic is systematic to an extreme. And this division
in feeling corresponds to the historic division. For iddje-, brahte-,
munde-, etc. , are the aorist stems in e-.
In WG. and NG. the weak preterit substituted the perfect
endings of the plural for the e- aorist endings. So Germ. *sagem^,
sage'di, sagen became in ON. *sggom, sggciS, *sggo (cf. OSw. laghu).
Words of this type were no doubt common at an earlier time, but
they were more and more displaced by the d- preterits. The appear-
ance of these is best explained by the Wackernagel-Behaghel theory
that the ending -thes of the 2d sing. (Skt. -thas, Gr. -6r]s, Olr. -the)
gave a feeling for a new tense stem. Thus pre-Germ. *sagdhes
(from *saghthes) gave a lengthened stem *sagdhe- beside *saghe-.
Later still arose the analogical preterits with an intermediate vowel:
Goth, lagida, salboda, habaida, OHG. legita, salbota, habeta. In part
these must have arisen in pre-Germ. The new tense stem thus
made would end in -dhe-. For -dhe- would naturally be abstracted
from *sagdhe-, *libdhe-, etc., to form Hoghidhe-. This -dhe- might
1 This would phonetically be -in with shortening of e before the nasal combination.
But the long vowel would be carried through by analogy as in Goth, salbond, OHG. sal-
bont, habent, etc.
126
Some Verb-Fohms in Gekmanic 63
have been associated in feeling with the root *dhe-, dho- 'set, do/
but there is no good reason for assuming that the weak preterit had
any other connection with the root *dhe-.
This explanation does not exclude the possibility that in some
instances the weak preterit was formed on a stem ending in t. Thus
to to- participles were formed te- aorists. In some cases also the t
was a part of the verb stem, as: OHG. fermisson:missen; ON. olla:
valda.
As an original e- aorist may also be explained OSw. byggi, ONorw.
biuggi, biugge 'wohnte,' bioggio, buggio 'wohnten.' Compare Gr.
^wj 'grew,' Lat. fue-re 'were,' OBulg. be 'was,' and the analogical
pret. Goth, bauaida 'wohnte.'
If the weak preterit is an e- aorist the 1st sing, would end in
-em. This would give Goth, -a as in munda, bauhta, nasida. The
-a in ON. is for older -o:Run. tawido. This -5 according to Brug-
mann, Grdr. IP, 3, p. 457, may represent IE. ou as in Skt. dadhdu.
But since WG. has -a: OHG. teta, nerita, OS. deda, nerida, OE.
dyde, nerede, which cannot represent IE. -ou (cf. Skt. aQtdu, Goth.
ahtau, OHG., OS. ahto, OE. eahta), it is more probable that the -a
in ON. is the same in origin as in WG., namely from -om, and that
this -dm came from the 1st sing. *dhedh6in 'did,' *ses6m 'sowed,'
etc. The usual interchange between the -o of the 1st sing, and the
-e- in the 2d and 3d may also have had something to do with this
change.
The 3d sing, would end in -et in the e- aorist, from which regularly
Goth, -a in Goth, munda, bauhta, etc., -e, -i in ON. hafe 'hatte,'
skulde, suafSe; e in OE. lifde; and -e in OS. and OHG. for the usual
-a. After the analogy of these is conjugated the 3d sing, in ON.
sere, rere, etc., and OE. dyde. The -a in OS. and OHG. is probably
regular in deda, teta {*dhedh6t) but analogical in thdhta, dahta {*tonk-
tet). The assimilated forms in -o as in OHG. ruarto mo (Braune
Gram. § 319, Anm. 1) may indicate the original -o in teta.
That pre-Germ. *dhedhdm would give the forms that occur admits
of no doubt. But whether *dhedhdt would result in OS. deda, OHG.
teta is not altogether certain. For Germ, we may safely assume
*dedd[J>]. Inasmuch as this o was protected, it might well have
developed differently from final unprotected o, which becomes
127
64 Francis A. Wood
WG. -u. The only parallel for final -ot occurs in noun stems in
-ot as Goth, mena, ON. mane, OE. mona, OS., OHG. mano 'Mond':
Goth, menops, etc. But here analogy has dictated the forms.
Evidence for the e- aorist may also be seen in the pret. opt. of
weak verbs. Only from an unthematic stem ending in a consonant
or from an e- aorist could the opt. have -1-. So OHG. missm 'miss-
ten,' wissis 'wiisstest,' muom,n 'miissten,' brdhtit 'brachtet,' neritit,
salbotlt, habetit, etc., may be compared in formation with Gr.
Trna^f tre, Seix^elre, Ti.nr)dtlT€, etc.
In the other WG. dialects and in NG. occur corresponding forms.
In Goth., however, are found only the longer forms made from the
stem of the plur. ind. Thus ind. wissed-un, kunped-un, mahted-um,
nasided-un, etc.: opt. mssed-eis, kunped-eip, mahted-i, nasided-i.
Francis A. Wood
University of Chicago
128